From: Xuxon | #2206 It's non-linear, dammit! Transitivity has nothing to do with percentages, it's the property where if X > Y and Y > Z, X > Z.
Yeah, that bugs the crap out of me as well! People always say non-transitive when they mean non-linear, even though I've been pointing this out for years :o(
Non-transitive pretty much doesn't exist as far as I can recall, other than L-Block in the 4-ways when he lost to one character one round but beat them the next.
Maybe Starcraft was non-transitive, but only very slightly, and only due to outright cheating.
Starcraft wasn't non-transitive, that would imply Melee loses to Halo, KH, or WW. No way.
There are many 4-way examples of simply reversing results, like Gordon/Scorpion and Alucard/Liquid, but given the absence of a third character even that wouldn't be non-transitive right? L-Block has some 3 character cycles but I can't think of any others off the top of my head.
Advokaiser | Posted 11/26/2011 8:44:30 PM | message detail What should have been the trend then?
Both Sora and Hitler acted like normal characters. It's tough to know who really won the board vote, as the rallying kicked in after 12:15 PM. Hitler only seemed to have a slight edge over Sora with the board vote, and that's despite the fact that Sora also gets anti-voted at the start of a match. And you figure Hitler probably had some anti-votes as well.
-- Luster Soldier --- ~Shield Bearer~ | ~Data Analyst~ Popular at school, but not as cool as BlAcK TuRtLe, Guru Champ!
Please stop conceding points to me only to claw them back over the next 1-2 matches.
You have done it virtually every time I've thought 'Well that gap is just about getting to uncatchable levels without upsets'. It is giving me false hope.
Either pull away or let me catch up. Pick one!
Many thanks,
SE
--
Crucifying my Oracle average one match at a time
ShatteredElysium, an 18 point gap isn't really that big. There's 20 matches to go. Actually a lot of people are doing well this year, and lots are within realistic distance.
I just noticed that I passed Dilated Chemist for #1 on the most #1 picks list! I now have 15. Cool.
Also if Ngamer finishes in the top 5 leaderboard this time, that would be his *4th* straight top 5 finish! Extremely impressive. Great to have such a badass partner.
Team PD is only 7 points back in the team race...lefty and gitanil should be quaking in fear :P We are like Jaws, always lurking...
Oh I know. It's sort of on the cusp of being chaseable. I think any higher and it would require mistakes from Lefty for me personally to do it. That's not saying others wouldn't be able to chase it though cos I'm certainly nowhere near the best Oracle predictor!
And I think he means overall Top 5's rather than individual matches Red Sox. You're projected to come second Ngamer and then win next year! 5th > 4th > 3rd previous contests so it would be natural progression!
--
Crucifying my Oracle average one match at a time
yaay I'm gonna be back in the top 20 after this match! Now I must stay in there until the end of the contest so I can get my 8th top 20 finish in a row
I'd like to make it back to single digits but that just doesn't seem possible at this point. Doesn't help that I haven't put any thought into a prediction in about two weeks. Would have liked my 10th top 10.
Come on battle bracket!
--
Rusty Bicycle: Rust is the perfect brake. You are technically correct, the best kind of correct!
Whoops, I misread the user profile page - if Ngamer finishes top 5 it would be his 3rd straight, not 4th straight. Still very impressive though. I've never managed more than 2 straight. Hopefully he wins this year - he's definitely the best predictor not to win. His points above average is 4th, although essentially it's identical to mine (1.13 vs. 1.07).
Carvey - can you check something for me? My predictions page said my prediction for match #36 (Sub-Zero vs. Scorpion, Donkey Kong vs. King K. Rool) is 69.16%. But my recorded prediction on my own Excel sheet is 66%. This must mean I made a late change but didn't record it. Can you just check on it for me though? The topic with that prediction has been purged it seems.
BTW, are the old prediction topics uploaded anywhere?
No prediction topics have been uploaded since the first contest in 2010. And even then, many of them are missing. Back when the topics used to be uploaded, I don't know if the archiving process was done automatically or if someone had to manually upload them.
-- Luster Soldier --- ~Shield Bearer~ | ~Data Analyst~ Popular at school, but not as cool as BlAcK TuRtLe, Guru Champ!
Mario with 68% says that Kirby / Meta would only 60-40 Ike / BK. 1v1 x-stats have that as a doubling ... before Meta Knight contributes more than Black Knight.
Either that or 68% is banking on 8-10% of SFF.
--
"Ovelia ... SPOILERS IN THE KEYWORDS WOOPS" - Vormav
That's a little too much SFF for me to think that's possible. Very few people are going above 65% here. It seems to be mostly concentrated in the 58-63% range.
-- Luster Soldier --- ~Shield Bearer~ | ~Data Analyst~ Popular at school, but not as cool as BlAcK TuRtLe, Guru Champ!
I'm not sure why people were going that low with Mario/Bowser... I'd expect Mario to 65-35 Sonic, plus Mario's rivalry is way more classic.... oh well, good for me (my safety pick was 68.78% though auuuugh)
there are two users with 100+ leaderboard appearances who have never spent a single day in first ...and they just so happen to be our current 2nd and 3rd place users
and thanks to my collapsing, they're now both within striking distance >_>
-- thelefty for analysis crew 2008 imo -tranny I have a third degree burn in flame-o-nomics -Sir Chris
I definitely have to play from behind if I want a chance to win. I really blew it today when I raised my prediction to 58.73% instead of staying at my original prediction of 55.73%. I was panicking when I saw everyone going near 60% on this match, so I foolishly raised my own prediction.
I could have gained nearly 4 points on lefty with my original 55.73%, but now I'm only going to gain slightly less than 1 point on him.
-- Luster Soldier --- ~Shield Bearer~ | ~Data Analyst~ Popular at school, but not as cool as BlAcK TuRtLe, Guru Champ!
In this contest, I've been making more of an effort to stay with my original prediction and it's worked out pretty well for me. The only time I've been making big changes to my predictions is when I realize my original prediction sucks big time. My Link/Ganondorf prediction for this round really sucks after seeing what Mario/Bowser just did yesterday, but I haven't gotten around to raising that one yet.
I have been considering going for some upsets in this contest in an attempt to gain an advantage over the users above me, but I haven't followed through with any upsets yet. I considered Dante/Vergil as an upset option, but I knew it just wasn't the right match to go for the upset. I even got a Top 5 on that Dante/Sora match for not even changing my original prediction. Trainers/Fighters is another match I'd consider the upset on, but I just feel that match isn't the right one to go for the upset. From the looks of it, the Trainers are a bigger favorite over the Fighters than Dante was against Sora.
-- Luster Soldier --- ~Shield Bearer~ | ~Data Analyst~ Popular at school, but not as cool as BlAcK TuRtLe, Guru Champ!
I'm tempted to change my oracle to Alucard / Dracula - 58.52% just for the lulz of mirroring 2010 Alucard > Magus. Wouldn't be the biggest curveball of the contest; Frog / Magus against Chell / GlaDOS wasn't exactly impressive either!
--
"Ovelia ... SPOILERS IN THE KEYWORDS WOOPS" - Vormav
I'd be shocked if Link/Ganon didn't surpass Mario/Bowser's percentage. I don't think Mega Man/Wily is THAT much stronger than Sonic/Robotnik. They certainly didn't look it against Zidane/Kuja.
--
"The great GF...Bahamut." "...GF? I...? Using my powers...It is you humans...I fear..."
Well I think the real question is not whether MM/Wily is as strong as Sonic/Robotnik (I'd consider them about equal), about obviously Link is stronger than Mario, it's whether Link can get the level of SFF that Mario got on Sonic. Mario/Sonic probably has more overlap than Link/MM, although we've seen Link SFF MM before (worth noting the pic was highly biased in that one though).
Your thoughts came up a couple of times on the new episode of The Show. Here's hoping you have a chance to give it a listen and perhaps even leave some comments for our listeners in the topic!